This paper uses a real options approach to examine the impact of abrupt increases in carbon dioxide emissions and pollutant-related socio-economic costs. It derives optimal investment rules in the form of critical values for both pollutant stock levels and social costs, above which environmental policies should be adopted. Moreover, it determines the optimal emissions abatement level. Our analysis extends the methodology of Pindyck (2000) using jump diffusion processes. We show that if the stock of pollutant is subject to extreme variations and the emissions abatement level is chosen exogenously by the policymaker, then lower levels of the pollutant stock are required to trigger policy adoption. A similar, yet more prominent, effect is observed under the assumption that pollutant-related socio-economic costs and benefits are expected to exhibit abrupt changes. However, different results are obtained when we examine simultaneously the two interrelated decisions, namely, the optimal threshold of emissions abatement and the optimal abatement level. In this case, an increase in the size and/or probability of a jump increases the critical values of both pollutant stock levels and socio-economic costs but leads to higher optimal abatement.